Many countries are installing photovoltaic (PV) systems as part of a shift towards a low carbon economy. This thesis describes the development of new PV system monitoring techniques to improve understanding in PV system performance. One hundred and nine domestic PV systems are studied and their energy performance is revealed through high resolution monitored data. The PV systems studied were installed from 2002 to 2004 as part of the UK Government's PV Domestic Field Trial. Located in five clusters throughout the UK, the sites are a combination of new build and retrofit projects. Each house has a separate grid-connected PV system and the 109 systems cover a variety of orientations, PV array sizes, PV module types and inverter sizes. The monitored performance data includes measurements of in-plane solar irradiance, PV module temperature, and DC and AC energy outputs. Data is recorded as five minutely average values for periods of up to two years. Data loggers store and provide access to the data through the public telephone network. New techniques have been developed to analyse the five minutely monitored data from each house directly without averaging or aggregating the values. Data quality and plausibility checks are undertaken to ensure the reliability of the measurements and the monitoring techniques are developed using data from fifty four of the PV systems. Plots of five minutely irradiance versus efficiency values show clouds of points which represent 'normal operation'. A statistical method is developed to detect faults and to separate them from the normal operation. Six types of normal operation energy losses and four types of faults are identified and these are investigated through original techniques including a method of constructing 'efficiency curves' for the different stages of normal operation and a method to detect shading based on the position of the sun. The reduction in overall annual performance caused by each type of loss and fault is calculated and the opportunities to minimise the losses and faults are investigated. This approach demonstrates a new strategy for monitoring PV systems, based on directly analysing high resolution monitored data. The method provides detailed insight into PV system performance and the results can be used to make recommendations for raising the efficiency of existing and future PV systems
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:435226 |
| Date | January 2006 |
| Creators | Firth, Steven |
| Publisher | De Montfort University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/2086/4102 |
Page generated in 0.0019 seconds